what is the different between domains" general (compact) " and "general (compact) s1"

[oo2316oo]

we see there are two different domain with similar name. would we know the difference?

---

Document Details

⚠ Do not edit this section. It is required for docs.microsoft.com ➟ GitHub issue linking.

• ID: 916de6f5-ef41-c82b-5918-0c6149326aab
• Version Independent ID: c7a4bb67-957d-f0a3-4a9f-e71035923f58
• Content: Export your model to mobile - Custom Vision Service - Azure Cognitive Services
• Content Source: articles/cognitive-services/Custom-Vision-Service/export-your-model.md
• Service: cognitive-services
• Sub-service: custom-vision
• GitHub Login: @anrothMSFT
• Microsoft Alias: anroth

[RohitMungi-MSFT]

@oo2316oo Thanks for the feedback. We are investigating the issue and will update you shortly.

[RohitMungi-MSFT]

@oo2316oo General (compact) [S1] is a new model architecture and is smaller than the current compact general model so can run on smaller devices. You can still export this in multiple formats, it is just not compatible with the Vision AI Dev Kit, so you don't have to select this option up front, you will select it when you go to export after the model is trained.

Products on shelves is a new domain model targeted at dense object detection and classification, particularly retail shelves, though could be used in other similar settings.

Our team is working to update the documentation with more details about the new domain types which will be updated soon.

[RohitMungi-MSFT]

@oo2316oo We will now proceed to close this thread. If there are further questions regarding this matter, please tag @RohitMungi-MSFT in your reply.

[Chonwing]

Hello, any idea about how many classes(objects) is the pre-trained model that is provided on the site where you can upload an image have. how many objects is it trained on and what are they? and can I add more classes to them. I read in the documentations that there thousands of classes. Is there more documentation about what the classes are?

[Oghost]

Hello @RohitMungi-MSFT Where I can find more documentation about the pre-processing to run in onnxruntime CustomVison General S1 OnnxModel's?

I am missing some normalization? if some one can oriented me I will appreciate.

actual code:

var size = new Size(300, 300); using var bmpResized = BitmapTransform.Resize(image, size); using var clone = image.Clone(new Rectangle(new Point(0, 0), size), System.Drawing.Imaging.PixelFormat.Format8bppIndexed); int width = clone.Width; int height = clone.Height; var inputMeta = _session.InputMetadata; var name = inputMeta.Keys.ToArray()[0];

        // pre-processing
        var dimentions = new int[] { 1, 3, height, width };
        var tensors = image.ToFloatTensor(false);
        var inputData = tensors.Merge(true);

        // session run
        var t = new DenseTensor<float>(inputData, dimentions);
        var inputs = new List<NamedOnnxValue> { NamedOnnxValue.CreateFromTensor(name, t) };
        var results = _session.Run(inputs).ToArray();
        var length = results.Length;
        var confidences = results[length - 1].AsTensor<float>().ToArray();